Abstract
BACKGROUND: We present a new methylation-sensitive amplified polymorphism (MSAP) approach for the evaluation of relative quantitative characteristics such as demethylation, de novo methylation, and preservation of methylation status of CCGG sequences, which are recognized by the isoschizomers HpaII and MspI. We applied the technique to analyze aluminum (Al)-tolerant and non-tolerant control and Al-stressed inbred triticale lines. The approach is based on detailed analysis of events affecting HpaII and MspI restriction sites in control and stressed samples, and takes advantage of molecular marker profiles generated by EcoRI/HpaII and EcoRI/MspI MSAP platforms. METHODS: Five Al-tolerant and five non-tolerant triticale lines were exposed to aluminum stress using the physiologicaltest. Total genomic DNA was isolated from root tips of all tolerant and non-tolerant lines before and after Al stress following metAFLP and MSAP approaches. Based on codes reflecting events affecting cytosines within a given restriction site recognized by HpaII and MspI in control and stressed samples demethylation (DM), de novo methylation (DNM), preservation of methylated sites (MSP), and preservation of nonmethylatedsites (NMSP) were evaluated. MSAP profiles were used for Agglomerative hierarchicalclustering (AHC) based on Squared Euclidean distance and Ward's Agglomeration method whereas MSAP characteristics for ANOVA. RESULTS: Relative quantitative MSAP analysis revealed that both Al-tolerant and non-tolerant triticale lines subjected to Al stress underwent demethylation, with demethylation of CG predominating over CHG. The rate of de novo methylation in the CG context was ~3-fold lower than demethylation, whereas de novo methylation of CHG was observed only in Al-tolerant lines. CONCLUSIONS: Our relative quantitative MSAP approach, based on methylation events affecting cytosines within HpaII-MspI recognition sequences, was capable of quantifying de novo methylation, demethylation, methylation, and non-methylated status in control and stressed Al-tolerant and non-tolerant triticale inbred lines. The method could also be used to analyze methylation events affecting CG and CHG contexts, which were differentially methylated under Al stress. We cannot exclude that the methylation changes revealed among lines as well as between Al-tolerant and non-tolerant groups of lines were due to some experimental errors or that the number of lines was too small for ANOVA to prove the influence of Al stress. Nevertheless, we suspect that Al tolerance in triticale could be partly regulated by epigenetic factors acting at the level of DNA methylation. This method provides a valuable tool for studies of abiotic stresses in plants.